On 2 August 2016 at 19:07, Eric Auger <[email protected]> wrote:
> From: Pavel Fedin <[email protected]>
>
> Introduce global kvm_arm_msi_use_devid flag and pass device IDs in
> kvm_arch_fixup_msi_route(). Device IDs are required by the ITS.
>
> Signed-off-by: Pavel Fedin <[email protected]>
> Signed-off-by: Eric Auger <[email protected]>
>
> ---
>
> v3 -> v4:
> - OR route->flags with KVM_MSI_VALID_DEVID
> ---
> target-arm/kvm.c | 6 ++++++
> target-arm/kvm_arm.h | 3 +++
> 2 files changed, 9 insertions(+)
>
> diff --git a/target-arm/kvm.c b/target-arm/kvm.c
> index dbe393c..4675aa3 100644
> --- a/target-arm/kvm.c
> +++ b/target-arm/kvm.c
> @@ -22,6 +22,7 @@
> #include "cpu.h"
> #include "internals.h"
> #include "hw/arm/arm.h"
> +#include "hw/pci/pci.h"
> #include "exec/memattrs.h"
> #include "hw/boards.h"
> #include "qemu/log.h"
> @@ -31,6 +32,7 @@ const KVMCapabilityInfo kvm_arch_required_capabilities[] = {
> };
>
> static bool cap_has_mp_state;
> +bool kvm_arm_msi_use_devid;
>
> int kvm_arm_vcpu_init(CPUState *cs)
> {
> @@ -619,6 +621,10 @@ int kvm_arm_vgic_probe(void)
> int kvm_arch_fixup_msi_route(struct kvm_irq_routing_entry *route,
> uint64_t address, uint32_t data, PCIDevice *dev)
> {
> + if (kvm_arm_msi_use_devid) {
> + route->flags |= KVM_MSI_VALID_DEVID;
> + route->u.msi.devid = pci_requester_id(dev);
> + }
> return 0;
> }
>
> diff --git a/target-arm/kvm_arm.h b/target-arm/kvm_arm.h
> index 633d088..befcfd1 100644
> --- a/target-arm/kvm_arm.h
> +++ b/target-arm/kvm_arm.h
> @@ -121,6 +121,9 @@ bool write_kvmstate_to_list(ARMCPU *cpu);
> void kvm_arm_reset_vcpu(ARMCPU *cpu);
>
> #ifdef CONFIG_KVM
> +
> +extern bool kvm_arm_msi_use_devid;
> +
The kernel documentation and most of the code for this flag
isn't ARM specific, which suggests to me that we should put
the QEMU support into the non-architecture-specific files too.
That is, a flag and wrapper function like all the existing
ones in include/sysemu/kvm.h, and then put the
if (kvm_msi_use_devid()) {
...
}
in the functions in kvm-all.c that need it.
thanks
-- PMM
